368 REPORT OF THE HORTICULTURAL DEPARTMENT OF THE 
TABLE I:— CHECK CASTRATIONS. 



Number of | Number of | Average’ 
NAME OF VINE. elusters | fruits fruits per 
castrated. developed. cluster. 
Gillett ned. seedling: } . olarak .OF. 1iaT. YA + 20 dana 
CACO RS. 8 hic re cae ne tds ha ee enn 8 0 
Beacons... oes ey iee ati nee eee 1 5 0.56 
ACA oN, Oe Be Ce ee pee Ces, See: 1 0 0. 
Concord D Rnien es SOTENS eas’ etal sh aie eae hake eh CERT E Le ener atten 19 0 ) 0 15 
Coucord: .ootimibe. tov awiphoddons. eh 1 34 ve 
Disiiong a0. <a cart ee chet 0 ane RE Ee tod 15 0 
Diambhid (24%) CMAG ROH eit Lee, 7 6 0, a¢° 
bl AFL LOT 4 Bick on anaienees ce ate mike tests oe 1 0 0.. 
INIGZEIS Ce Te eS eS ee ert e er 13 0 0. 
Pocklmegtoni+. fun. tt 20) 10. tone 6 0) 0 
Wroruen. rere tore eee. phe ga 12 8 f rate 
Worden SAT EE RTI OR | MAS ALA as 4 8 0.5 
CRO gt wet dali bank th Rete 4 be : 86 42 Sa 


Taking all clusters into account, the average number of fruits 
produced per cluster was 0.49. Estimating the number of cas- 
trated blossoms per cluster at 50, it appears that less than 1 per 
ct. of the castrated blossoms developed into fruit. The method 
of castrating the grape with thumb and finger, if used with care, 
is well adapted for an experiment like the one under considera- 
tion. In this case so many blossoms were included in the tests 
that the general results may be accepted as giving a reliable 
indication of the potency of the pollen tested. In considering the 
results of the experiments, the fact should not be lost sight of 
that nearly 1 per ct. of the check castrations developed ols 
from accidental self-pollination. 
After castration the clusters were iriiediatete inclosed in 
paper bags, which were closed with a wired label. This cover- 
ing was not removed again till the operation of cross-pollination 
was undertaken. 
Pollination.—When the varieties which had been chosen to fur- 
nish the supply of pollen came into bloom the covered clusters 
were cut off and, without removing them from the bag, were 
taken to the castrated clusters which were to be eross-pollin- 
ated. The stigmatic surfaces of the latter were then touched 
